Friday, March 25, 2022

Grim Dawn's Dungeon Blockers and Ye Olde Evil Doors

Grim Dawn does a couple of things in its dungeon that are sometimes annoying but fundamentally clever.  There are no randomly-generated maps in Grim Dawn.  The closest thing is a randomly-generated infinite dungeon, but each level is drawn from a pool of fixed level maps.  Given that a typical character runs through all of the campaign content in the game several times, in the traditional Diablo model, one would expect players to develop highly-efficient routes through the campaign area maps.

In order to prevent these repeatedly-traversed campaign zones from becoming stale without having to resort to random generation (with the difficulties that that would entail, like nonsensical outputs), the Grim Dawn devs have put places on some maps (mostly dungeons and towns, but a few wilderness areas) where random blockers can appear.  Piled of burning garbage, doorways that are full of rocks, that sort of thing.  They have the potential to appear in a fairly small number of carefully-selected spots in each dungeon, they're re-randomized per session (save and reload, basically, which also returns you to town), and generally you can't see them until you get fairly close to the spot they're blocking.  What all this means is that you can't consistently run optimal routes that you plan out before entering the dungeon.  There's a good chance that any such planned route will be blocked and you'll have to detour off it and then find a way back onto it, and your detour might also be blocked!  This means that finding paths remains gameplay even with hundreds of hours in the game and tens of runs through each area.  There are real choices and problem-solving in path-selection even when you have perfect knowledge of the unrandomized parts of the underlying level.  So despite having a constant underlying structure, navigating these areas remains engaging.

This seems like something which, obviously, could be adapted to OSR dungeons.  As DMs, we're under the same pressures to reuse content that Grim Dawn's devs were, and fully-procedural dungeons carry similar difficulties.  And it would be consistent with the No Homework ethos - no homework'd plan to traverse known parts of the dungeon will necessarily survive.  Reducing the rewards for homework is likely to reduce the amount of homework done.

Another clever, engaging thing that Grim Dawn does in its dungeons is find any excuse to have enemies "appear" in close proximity or behind you while you've engaged enemies in front of you.  Zombies, skeletons, and giant insects burrow out of the ground behind or around you.  Spiders drop from the ceiling.  Wraiths and outsiders materialize out of thin air.  Often these animations take long enough that if you're in motion, you can leave these appearing monsters in the dust and never have to deal with them.  This works great if you're running a nice planned path...  but when you run into a blocker and have to backtrack, or run into stern opposition, these guys can really catch up with you.  So blowing past them is a gamble, and they serve as a sort of potential energy mechanic.

But obviously, it would be silly to have goblins just appear out of nowhere near and behind you.  Unless...

Enter the much maligned stuck/evil door.  Obviously, goblins can come out of it on short notice!  And where better to place planned-path-breaking blockers than in doorways?  You could place your careful cut-points in hallways, but your dungeon already has all these doorways!  But it's not quite the same, because they're not blocked per-expedition, and a party can retry opening them.  Right?

Well...  not necessarily.  Looking at OD&D and the 1e DMG, it really isn't clear how long it takes to open a stuck/evil door, and doesn't explicitly say that it can be retried.  Not being able to retry forcing would give you a good reason to bring out the axes and exercise the door-breaking mechanic, so in a certain light the door-breaking mechanic's existence might be circumstantial evidence for rejecting retries on forcing doors.  And indeed there are apparently folks who play without permitting retries to open stuck doors - this post from Knight at the Opera, down in the "Dense Megadungeons" section, takes this position, and examines the effects on potential paths through the dungeon.  A door which the party failed to open on the first d6 roll was stuck closed to them for the rest of that expedition into the dungeon.  Knight at the Opera mapped out the paths that four expeditions had taken as a result of doors that they couldn't open.  Their patterns of exploration ended up deeper and more linear in structure than the patterns of dungeon exploration I'm used to seeing, because if they found a room with three doors, probably only one of them would open, so they explored where they were able, subject to door luck.  And obviously a door amenable to being opened by the party could then also be reopened within the same expedition (though it might require a roll that could be retried).

It's still not quite the same as Grim Dawn's dungeon blockers, but if you change the probabilities to something like 1 in 6 doors stuck (for the party) in any given expedition, I think you'd end up with a similar disruption of planned paths through explored areas and a similar excuse to have monsters sneak in behind parties, without wildly altering the pattern of exploration.  One complication is that you may want to leave all areas accessible during any given expedition, which could require sometimes rejecting a roll's result (if a stuck result would block the last path to an area).  But that's a rather principled sort of rejection of a random result, and one with good gameplay consequences (probably?).


  1. If you can just break down the doors, what's the point of them being stuck? Just to force a wandering monster check when you break it down?

  2. Breaking down a door took three *turns* in pre-Wizards of the Coast D&D, which is enough time for a plate-armored party to travel 18 10' squares at exploration speed, or for a lightly-armored party to travel up to 36 10' squares. It's very possible for alternate routes to be quicker than breaking down a door, without incurring any additional random encounters from noise. Just the time to break down a door alone guarantees a random encounter roll from time elapsed, with a possibility of a second as well, before any from noise. It also consumes half a torch and a good chunk of the duration of many spells.

  3. I've never heard or played with the "three turns" figure before, and it doesn't seem to be in the 1e DMG - where does it come from? I guess it could have been based on this:

    DMG pg 97, The Campaign (The First Dungeon Adventure) > Doors:

    ... If wooden doors (always metal bound, naturally) are broken down by axes and the like, it will take some time - a full turn is usual - and require at least 3 checks to see if nearby and/or wandering monsters are attracted by the noise. ...

    1. Huh, you're right - I had read that section of the 1e DMG last week and must have misinterpreted/misremembered it. Now I'm wondering where ACKS got its three turns to batter down doors rule.